Saving lives

through research
and education

Hidden Highways:
Fog and Traffic Crashes on
America’s Roads
November 2014

607 14th Street, NW, Suite 201 | Washington, DC 20005 | AAAFoundation.org | 202-638-5944
Title

Hidden Highways: Fog and Traffic Crashes on America’s Roads (November 2014)

Author

Bruce Hamilton
Brian Tefft
Lindsay Arnold
Jurek Grabowski
AAA Foundation for Traffic Safety

About the Sponsor

AAA Foundation for Traffic Safety

607 14th Street, NW, Suite 201
Washington, DC 20005
202-638-5944
www.aaafoundation.org

Founded in 1947, the AAA Foundation in Washington, D.C. is a not-for-profit, publicly

supported charitable research and education organization dedicated to saving lives by
preventing traffic crashes and reducing injuries when crashes occur. Funding for this report
was provided by voluntary contributions from AAA/CAA and their affiliated motor clubs,
from individual members, from AAA-affiliated insurance companies, as well as from other
organizations or sources.

This publication is distributed by the AAA Foundation for Traffic Safety at no charge, as a
public service. It may not be resold or used for commercial purposes without the explicit
permission of the Foundation. It may, however, be copied in whole or in part and
distributed for free via any medium, provided the AAA Foundation is given appropriate
credit as the source of the material. The AAA Foundation for Traffic Safety assumes no
liability for the use or misuse of any information, opinions, findings, conclusions, or
recommendations contained in this report.

If trade or manufacturer’s names are mentioned, it is only because they are considered
essential to the object of this report and their mention should not be construed as an
endorsement. The AAA Foundation for Traffic Safety does not endorse products or
manufacturers.

©2014 AAA Foundation for Traffic Safety

Abstract

Although fog and smoke are understood to create challenging driving conditions for
motorists, surprisingly little research has been conducted on the characteristics of fog- and
smoke-related crashes, and on the prevalence of such crashes in overall national highway
safety statistics. This report illustrates the scope of the problem by presenting 23 years of
national data on fatal crashes involving fog and smoke, and 19 years of police-reported
crash data pertaining to these conditions. These data point to issues of particular concern
with regards to fog and smoke, including the elevated prevalence of such crashes among
young drivers, during winter months, and along undivided rural highways. Perhaps most
alarmingly, fog and smoke appear to play a major role in fatal multi-vehicle pileups, and
are coded as a factor in nearly one-in-five such crashes involving 10 or more vehicles. By
examining national and regional data, as well as existing research into driver behavior and
perception changes due to fog, this paper offers recommendations that highway officials,
safety advocates, parents, automakers, driving instructors, and road users of all kinds can
act on in order to promote safe highway operations in foggy or smoky conditions.

2
Introduction

For three pre-dawn hours of January 29, 2012, the Florida Highway Patrol closed and
monitored a stretch of Interstate 75 near Gainesville, Florida, due to thick smoke and fog
that had contributed to a non-fatal three-car crash.1 As visibility improved, the roadway
was reopened, despite at least one Trooper’s insistence that conditions could change rapidly
and smoke and fog could return without warning.2 Less than an hour later, 11 people were
killed and 18 were injured in a 19-vehicle pileup attributed to a sudden drop in visibility
when smoke and fog again covered the freeway.1

Fog presents numerous challenges to motorists.1 By definition, it reduces visibility to less

than 5/8-mile,3 and, as fog obscures details of the driving environment and reduces contrast,
drivers have difficulty accurately perceiving certain things that are crucial for safe driving,
such as depth and speed.4 Moreover, because high beams reflect off of the suspended water
droplets that fog comprises and actually make it harder to see, drivers are deprived of a tool
that is useful in other low-visibility situations. And, as was seen in the Florida crash, fog
can blanket a roadway with little warning and create sudden and dramatic changes in
driving conditions.

Despite these dangers, research addressing the issue of fog and its impact on highway safety
is relatively sparse, though some illuminating studies do exist. In examining driver behavior
in fog, Broughton et al. determined that fog divides motorists into two camps, which they
called “laggers” and “non-laggers.”5 Non-laggers (the predominant group) – perhaps feeling
uncomfortable with the loss of visual reference points in the fog – try to stay within eyeshot of
the vehicles in front of them, even though doing so often requires following at an unsafe
distance.5 To make matters worse, motorists might not be aware of how dangerous their
chosen following distance is: Cavallo et al. found that their study participants perceived a
lead car to be 60 percent farther away in foggy conditions than in clear conditions.6 The
potential for crashes when drivers simultaneously try to maintain visual contact with lead
vehicles and fail to realize how close they have to come to do so is evident.

Safe following distances are, of course, related to travel speeds, and here too the existing
research highlights some issues. Brooks et al., for example, found that drivers do not tend
to reduce speed in fog until their ability to stay in their travel lane is compromised.7
Because lane markings (which are close to the vehicle and often retroreflective) may remain
detectable even in certain limited-visibility conditions, Brooks et al. cautioned that drivers
can become overconfident in fog, and also found that when motorists do finally slow down,
the speed reduction is often not sufficient to avoid unexpected hazards.7 Furthermore,
Mueller et al. looked at this issue through the lens of driver experience, and found that
young novice drivers were the slowest to react to hazards, and reduced their speed the least
in response to foggy conditions.8

While these and other studies provide insight into driver behavior and perception of speed
and distance in fog, they do not address crash outcomes or trends related to foggy
conditions. Here the research is even sparser. Abdel-Aty et al., in an effort to fill this void,
took a detailed look at fog and smoke-related crashes in Florida, using crash data from

3
2003-2007.9 Among their findings were that fog- and smoke-related crashes were most
likely to occur during the winter months (December – February), in the early morning
hours, and in rural areas.9 They also found that crashes in smoke and fog tended to involve
more vehicles and more serious injuries than did crashes in clear visibility conditions.9

This study paints a more complete picture of fog- and smoke-related fatal crashes (FSFC)
and fog- and smoke-related police-reported crashes (FSPC) by examining national trends. It
offers a descriptive analysis of 23 years of crash data (1990-2012) from the National
Highway Traffic Safety Administration’s Fatality Analysis Reporting System (FARS), and
19 years of data (1990-2008) from the National Automotive Sampling System General
Estimates System (NASS GES). Such an analysis will help fill some of the knowledge gaps
regarding fog- and smoke-related crashes, and be of benefit to overall highway safety efforts
given the elevated potential for severe and multi-vehicle crashes in foggy conditions.

Methods

The data for this report come from two sources, both overseen by the National Highway
Traffic Safety Administration (NHTSA):

1) The Fatality Analysis Reporting System (FARS) is a national database that

documents every traffic crash on a public roadway in the United States involving at
least one fatality (within 30 days of the crash).
2) The National Automotive Sampling System General Estimates System (NASS GES)
provides national data from a representative sample of all types of police-reported
crashes, including fatal, property damage only, injury, etc.

Twenty-three years of FARS data (1990-2012) and 19 years of GES data (1990-2008) were
collected and analyzed for this report. Data were tabulated to track fog-related crashes as a
percentage of all fatal crashes (Table 1), and were weighted to reflect fog-related crashes as
a percentage of all police-reported crashes nationwide (Table 1b). Also of interest were the
characteristics of fatal crashes involving fog, including time of day, month, number of
vehicles involved, roadway alignment, etc. These are presented in Table 2. Additionally,
due to regional variations in fog prevalence, data were broken down by state and NHTSA
region (Table 3). Graphs were created from subsets of the data to highlight especially
noteworthy trends, and are presented throughout.

Complete data are presented in each of the tables; however, it is important to note that in
2007, the coding of the FARS “weather” variable changed. Through 2006, “fog” was a
separate and distinct category of the weather variable; beginning in 2007, “fog,” “smoke,”
and “smog” were all combined. As such, it is not possible to distinguish between these three
atmospheric events in the 2007-2012 data. For consistency, this report refers to “fog- or
smoke-related” crashes throughout, given that the effects that these conditions have on
visibility are largely similar. The data does distinguish between crashes involving the
coding of fog alone, and fog together with other types of precipitation, such as rain or snow.

4
Results

General Trends Over Time

Over the past two decades, the United States has experienced a general decreasing trend in
the number of fatal crashes involving fog (Table1; Figure 1), and the number of police-
reported crashes involving fog (Table 1b; Figure 2).

Annual Number of Fatal Crashes Involving Fog, United

States, 1990 - 2012
700

600

500

400

300

200

100

Fog Alone Fog + Other Precipitation

Figure 1. Graph compiled from FARS data (1990-2012 Final Files)

Annual Number of Police-Reported Crashes Involving Fog,

United States, 1990-2008
50,000

40,000

30,000

20,000

10,000

Fog Alone Fog

og ++Other
OtherPre
Precip
cipitation

Figure 2. Graph compiled from NASS GES data (1990-2008). Data weighted to reflect all
police-reported crashes nationwide

5
These trends are consistent with an overall reduction in both fatal and police-reported
crashes nationwide during this time. However, the good news is that there has also been a
decline in the percentage of overall crashes accounted for by fog- and smoke-related
incidents. This trend is shown below for fatal crashes (Table 1; Figure 3) and for all police-
reported crashes (Table 1b; Figure 4).

Fatal Crashes Involving Fog as a Percentage of All Fatal

Crashes, by Year, United States, 1990-2012
1.80
1.60
1.40
1.20
1.00
0.80
0.60
0.40
0.20
0.00

Fog Alone Fog + Other Precipitation

Figure 3. Graph compiled from FARS data (1990-2012 Final Files)

Police Reported Crashes Involving Fog, as Percentage of All

Police-Reported Crashes, by Year, United States, 1990-2008
0.80%
0.70%
0.60%
0.50%
0.40%
0.30%
0.20%
0.10%
0.00%

Fog Alone Fog++Other

Fog Other Precipitation
Precip

Figure 4. Graph compiled from NASS GES data (1990-2008). Data weighted to reflect all
police-reported crashes nationwide

6
Regional Variations

Not all parts of the country are equally susceptible to fog formation. Consequently, the
degree to which fog- and smoke-related crashes factor into overall crash numbers differs by
region and by state. Table 3 presents an overall state-by-state breakdown of fatal crashes
involving fog for the 23 years of data covered in this study. Figures 5 and 6 group and
present these data according to the 10 NHTSA regions to highlight broad variations across
the country.

Looking at numbers alone, Region 4 in the Southeast (AL, FL, GA, SC, TN), Region 5 in the
Midwest (IL, IN, MI, MN, OH, WI), and Region 6 in the South/Gulf Coast (LA, MS, NM,
OK, TX) experience the most fog- and smoke-related fatal crashes (Figure 5). The picture
changes, however, when looking at the proportion of fatal crashes in each region that
involve fog (Figure 6). Here the highest percentages are in Region 1 in New England (CT,
MA, ME, NH, RI, VT), Region 7 in the central U.S. (AR, IA, KS, MO, NE), and Region 10,
which includes the Pacific Northwest (AK, ID, MT, OR, WA).

Number of Fatal Crashes Involving Any Fog, by NHTSA Region,

1990-2012
2500

2000

1500

1000

500

Figure 5. Graph compiled from FARS data (1990-2012 Final Files)

Seasonal and Time-of-Day Patterns

Consistent with Abdel-Aty et al.’s findings in Florida, fatal crashes nationwide involving fog are
most prevalent during the winter months (December – February), when they account for 2.44
percent of all fatal crashes, and are least prevalent in the summer (June – August), when they
account for less than 1 percent of all fatal crashes (Figure 7). Additionally, the prevalence of
smoke- and fog-related fatal crashes is highest in the overnight (midnight – 5:59 AM) and

7
morning (6:00 – 11:59 AM) hours (Figure 8). This is not surprising, as the overnight and early
morning hours – when the temperature is generally the coolest and water vapors can condense
into droplets – are when fog formation is most likely to occur.10 Additionally, these are times
when visibility is generally already compromised due to poor lighting. Adding smoke or fog to
the mix further degrades already-challenging driving conditions.5

Fatal Crashes Involving Any Fog, as Percentage of All

Fatal Crashes by NHTSA Region*, 1990-2012
2.0%
1.8%
1.6%
1.4%
1.2%
1.0%
0.8%
0.6%
0.4%
0.2%
0.0%

Figure 6. Graph compiled from FARS data (1990-2012 Final Files)

*See Table 3 for state listing by NHTSA region.

Fatal Crashes Involving Any Fog, as Percentage

of All Fatal Crashes by Month, 1990-2012
3.0 %
N=185,926
2.5 %
2.0 %
N=218,791
1.5 %
N=199,789 N=227,681
1.0 %
0.5 %
0.0 %
Dec-Feb Mar-May Jun-Aug Sept-Nov

Figure 7. Graph compiled from FARS data (1990-2012 Final Files)

8
 Fatal Crashes Involving Any Fog, as Percentage
of All Fatal Crashes by Time of Day, 1990-2012
3.5 %
N=173,836
3.0 %
N=158,768
2.5 %
2.0 %
1.5 %
N=230,891
1.0 % ,926
N=242,555
0.5 %
0.0 %
Midnight-5:59am 6:00-11:59am 12:00-5:59pm 6:00-11:59pm

Figure 8. Graph compiled from FARS data (1990-2012 Final Files)

Land Use and Roadway Type

Variations in the prevalence of fog- and smoke-related fatal crashes are also seen when
looking at land use and roadway type. Prevalence is greater in rural areas – where such
crashes account for nearly two percent of fatal crashes – than in urban ones, where less
than one percent of fatal crashes are accounted for by fog (Figure 9). Additionally,
undivided, two-way roadways see a greater share of fog- and smoke-related crashes than do
divided or one-way roads (Figure 10). This national finding is consistent with what Abdel-
Aty et al. saw in the Florida data.9

Fatal Crashes Involving Fog, as a Percentage of

All Fatal Crashes by Land Use, 1990-2009
2.5%
N=423,474
2.0%

1.5%
N=314,313
1.0%

0.5%

0.0%
Rural Urban

Figure 9. Graph compiled from FARS data (1990-2008 Final Files; 2009 Annual File)

9
 Fatal Crashes Involving Fog, as a Percentage of
All Fatal Crashes by Roadway Type, 1990-2009
1.6% N=483,455
1.4%
1.2% N=174,085
N=59,309
1.0%
0.8%
N=8,639
0.6%
0.4%
0.2%
0.0%
Divided with Divided No Not Divided One-Way
Median Barrier Median Barrier

Figure 10. Graph compiled from FARS data (1990-2008 Final Files; 2009 Annual File)

Driver Age and Number of Vehicles Involved

The national data show that, in general, for drivers of older ages, fog- and smoke-related
fatal crashes account for a smaller share of fatal crash involvements. The proportion is
highest for drivers aged 20-29, for whom fog is coded in 1.54 percent of their fatal crashes,
and lowest for drivers aged 70 and older, for whom fog is coded in just 0.65 percent of fatal
crashes (Figure 11). Teens, at 1.46 percent, see the second highest proportion.

Another pattern seen in the national data with regard to fog- and smoke-related crashes
pertains to the number of vehicles involved. For single-vehicle fatal crashes, as well as fatal
crashes involving 2-5 vehicles, the prevalence of fog as a factor is relatively low and fairly
consistent, ranging from 1.40 percent to 1.47 percent. For crashes involving six to nine
vehicles, however, the prevalence more than triples, to 4.37 percent. And for pileups
involving 10 or more vehicles – such as the 19-vehicle disaster in Florida in January 2012
that claimed 11 lives – fog is coded in nearly one-in-five occurrences (Figure 12). While such
crashes are indeed rare (215 occurred over the 20-year study period, compared with more
than 482,000 fatal single-vehicle crashes), the alarming spike in prevalence suggests that
fog and smoke are indeed risk factors for this horrific and lethal crash type.

10
 Percent of Drivers in Fatal Crashes Involving Fog
(Relative to Drivers in All Fatal Crashes), by Age,
U.S. 1990-2009
2.00%

1.50%

1.00%

0.50%

0.00%
<20 20-29 30-39 40-49 50-59 60-69 70+
Age Group (years)

Figure 11. Graph compiled from FARS data (1990-2008 Final Files; 2009 Annual File)

Fatal Crashes Involving Fog, as a Percentage of

All Fatal Crashes by Number of Vehicles
Involved, 1990-2012
20.0 %

15.0 %

10.0 %

5.0 %

0.0 %
1 2 3-5 6-9 10+
Number of Vehicles Involved in Crash

Figure 12. Graph compiled from FARS data (1990-2012 Final Files)

Discussion

While fog- and smoke-related fatal crashes (FSFC) do not generally account for a large
percentage of overall fatal crashes (<2%), these numbers only tell part of the story. Whereas
crashes related to things like distraction, speeding, or impairment can happen at any time and
in any location, FSFC are dependent on the presence of specific atmospheric conditions to
occur. With some parts of the country – such as the Southwest – receiving very little fog, and
even relatively foggier places seeing little formation during the daylight hours and warmer

11
months, it is not surprising that the raw crash numbers are somewhat low. What is important
is that – given the presence of fog or smoke as opposed to clear conditions – crashes are more
likely to occur, to result in serious injuries, and to involve multiple vehicles.9 As such, highway
safety officials and other stakeholders should not discount the significant threat posed by fog
and smoke, and more research on the subject is in order, especially with respect to the
effectiveness of countermeasures that may be in use (e.g., fog detection systems).

The roughly two decades of national data presented here help to highlight some of the key
issues and concerns posed by FSFC. As shown in Figure 11, FSFC are more prevalent
among younger drivers, and are least prevalent among older ones (drivers 75+). This is
consistent with the fact that older drivers are among the most responsible on the road, and
tend to limit their driving at night and in inclement weather.12 At the same time, teens are
the most inexperienced and crash-prone drivers on the road, and may be unaccustomed to
dealing with the particular challenges posed by fog.11 In examining driver behavior changes
in fog, as noted earlier, Mueller et al. found that their young novice participants took the
longest to react to hazards and reduced their speed the least, and that one-in-four crashed
during the study.8 However, newly-licensed teens in virtually all states are restricted from
driving during the overnight and early morning hours – the times when fog formation and
related crashes are most likely to occur – which likely contributes to the slightly lower
prevalence of FSFC seen among this age group than among drivers a few years older.

The elevated prevalence of FSFC among younger drivers is yet another reminder of the
importance of giving novice drivers ample opportunities to practice their skills under a
variety of conditions, and existing AAA Foundation research indicates that there is
significant room for improvement in this area. In a naturalistic driving study of teens in the
learner stage of North Carolina’s graduated driver licensing (GDL) system, the Foundation
found that teens were getting less supervised driving practice than had been previously
assumed, and that most of the experience they did get was on bright sunny days and along
familiar routes.13 In fact, only 16 percent of the video clips showed teens practicing driving
in the dark, and just 3 percent were recorded in the rain.13 Providing opportunities for teens
to practice driving under a variety of conditions – including fog – while under the
supervision of parents or guardians is a key component of young driver safety efforts.

The role that fog and smoke appear to play in fatal multi-vehicle pileups (accounting for
nearly one-in-five fatal 10+ vehicle crashes) is particularly alarming, and indicative of the
myriad ways in which these conditions complicate a driver’s ability to perceive the road
environment. With reference points obscured and many motorists looking to lead vehicles to
provide some visual context, the potential for crashes rises, particularly since following
distances may need to be unsafely short in order to maintain visual contact.5 Indeed,
Hawkins – in a study from the United Kingdom – observed vehicles clustering in foggy
conditions, and the percentage of small gaps in traffic (under 60 meters) increasing by 25
percent.14 Such clustering has obvious implications for the risk of chain-reaction crashes
and multi-vehicle pileups, as drivers may not leave themselves enough time and space to
avoid colliding with vehicles around them. This may be particularly true given Cavallo et
al.’s findings that drivers perceive vehicles to be farther away in fog than they do under
clear conditions, and Broughton et al.’s findings that motorists’ differing approaches for
dealing with fog can interrupt the normal flow of traffic and yield discrepancies in speed
and vehicle positioning.6,5

12
Given these challenges, and the deadly risks they pose, greater public awareness of the
specific effects that fog and smoke can have on driver perceptions, speed choice, following
distance, and other behaviors should be promoted. Such efforts could take the form of
modules in driver education courses, seasonal public service announcement (PSA)
campaigns during “foggy times” (generally the winter and fall), and more detailed coverage
of the effects that fog and smoke have on driving when crashes related to these conditions
occur. Specific central themes of such messaging could include references to fog and smoke
making objects appear farther away than they are, interfering with driver ability to judge
speed, and leading to vehicle clusters on the highway.

While motorists and other road users certainly have a responsibility to take precautions
and exercise safe behaviors, highway officials and even vehicle manufacturers also have a
role to play in reducing the likelihood and severity of fog- and smoke-related crashes. The
breakdown of the national data into NHTSA regions (Figure 6) highlights areas of the
country where the implementation of countermeasures for fog- and smoke-related crashes
may be of particular benefit.

A number of roadway design features, for example, can assist motorists who are struggling
with diminished visual references and perception abilities. Rumble strips, for example,
provide a physical and audible warning – even when drivers cannot see the center or edge
line – that can alert motorists to an imminent roadway departure before it’s too late.
Upgraded pavement markings and plastic lane markers may help drivers maintain their
position even in thick fog, and widened shoulders can provide safe places for motorists to pull
over in the event they feel they can no longer drive safely given the conditions. Additionally,
median barriers – such as concrete walls, high-tension cables, or steel guardrails – can
prevent vehicles from crossing into oncoming traffic and causing severe head-on wrecks. The
national data on FSFC by roadway type (Figure 10) suggests such barriers may be
particularly beneficial countermeasures, as the prevalence of such crashes is highest on two-
way, undivided roadways. This is further supported in Abdel-Aty et al.’s data from Florida,
where they found that, while the likelihood of all crash types examined were higher in foggy
conditions than clear ones, the odds ratio was greatest for head-on crashes.9 If a barrier is
deemed infeasible, centerline rumble strips – which run down the middle of an undivided
roadway – can warn drivers that they are about to cross into oncoming traffic.

While highway officials can consider implementing crash countermeasures of the kinds
described above, there may be opportunities for reducing fog- and smoke-related crashes
through vehicle design and technology, as well. For example, in addition to replicating
Cavallo et al.’s findings that vehicles are perceived to be farther away in foggy conditions
than in clear ones, Buchner et al. demonstrated that subjects also perceived vehicles with
higher-positioned rear lights and rear lights that were closer together as farther away, and
suggested that such vehicles may be more likely to be rear-ended in fog.15 Rear lights that
are positioned closer to the ground, and with greater separation, may therefore help trailing
motorists better estimate and maintain a safer following distance.

Another issue that merits consideration here is the range of sophisticated crash-avoidance
technologies that are increasingly found in new vehicles. Recent research by the AAA
Foundation and the Massachusetts Institute of Technology has discovered that while some
of these technologies – such as Electronic Stability Control – have a well-documented safety
benefit, others still require substantial evaluation before they can be considered “proven.”17

13
For example, two of the technologies evaluated – lane departure warnings and forward
collision warnings – are intended to help drivers maintain proper vehicle positioning and
avoid rear-end crashes (two scenarios that are highly applicable to fog- and smoke-related
situations). However, in a consumer-facing rating system developed by the research team
(available at www.aaafoundation.org/ratings-vehicle-safety-technology), each of these
technologies received 3 out of 5 stars for their theoretical overall safety benefit, and just 1
out of 5 stars for what has been proven in real-world data to-date.17 And, in addition to this
general lack of data, more needs to be learned specifically about how foggy and smoky
conditions affect the sensors and/or cameras upon which these systems depend.

While the data presented here do not indicate why FSFC and fog- and smoke-related police
reported crashes (FSPC) have generally declined both in number and as percentages of overall
crashes over the past two decades, it is likely that several factors are at work. These may
include an increase in the number of vehicles with advanced technologies, roadway
infrastructure upgrades, and the enactment of state GDL laws restricting high-risk novice
drivers from operating motor vehicles during fog-prone overnight hours. It is also possible that
larger, climate-level changes have played a role. In examining fog trends across the
Southeastern United States (AL, FL, GA, NC, SC, VA) from the mid-20th century through 2003,
Forthun et al. found that nearly half of the measurement sites (12 of 26) experienced
significant reductions in fog-event days, and only two stations saw an increase.16 In addition,
medium- (3-5 hours) and long-duration (6+ hours) fog-events decreased everywhere that trends
were significant, and only five stations showed an increase in short-duration (1-2 hours) periods
of fog.16 Given that the national data show that the Southeast accounts for the greatest number
of FSFC (Figure 5), decreases in fog formation and duration in this region may have helped
drive down crashes related to these conditions. Forthun et al. note that they believe large-scale
and long-term meteorological shifts likely account for the significant decreases in fog-events,
but also acknowledge that others have attributed at least some of the decline to urbanization
and its associated increase in temperatures and decrease in humidity.16 While their study
looked only at the Southeast, it is certainly possible that the forces of urbanization or broader
climate change have impacted fog formation elsewhere, as well.

Conclusion

Despite the relatively low prevalence of fog- and smoke-related fatal and police-reported
crashes, and a general decreasing trend in such crashes over the past two decades, fog and
smoke remain significant threats to highway safety given the particularly insidious ways in
which these conditions appear to impact driver perceptions and behaviors. Given the
increased likelihood of crashes in the presence of fog and smoke – and, most troublingly, the
increase in severe and multi-vehicle crashes – fog and smoke should be treated as serious
safety concerns, and efforts should be made to continue developing and evaluating
countermeasures targeting the issue. The national data highlight certain priority concerns
with regards to fog – such as young driver crashes, winter driving, multi-vehicle pileups,
and undivided rural roadways – and, when considered alongside existing research into
driver behavior and perception, offer actionable ideas for highway officials, safety
advocates, parents, automakers, driving instructors, and other stakeholders wishing to see
reductions in crashes related to fog and smoke.

14
Tables

Table 1
Fatal Crashes by Police-Reported Fog Status, United States, 1990-2009
All Fatal
Fog Alone Fog + Other Precipitation Crashes
% of all fatal n % of all fatal
n (crashes) crashes (crashes) crashes N
1990 625 1.57 70 0.18 39,836
1991 583 1.58 61 0.17 36,937
1992 555 1.59 75 0.21 34,942
1993 505 1.41 44 0.12 35,780
1994 570 1.57 43 0.12 36,254
1995 488 1.31 43 0.12 37,241
1996 594 1.58 72 0.19 37,494
1997 503 1.35 45 0.12 37,324
1998 556 1.50 67 0.18 37,107
1999 461 1.24 48 0.13 37,140
2000 493 1.31 45 0.12 37,526
2001 548 1.45 41 0.11 37,862
2002 421 1.09 49 0.13 38,491
2003 474 1.23 51 0.13 38,477
2004 493 1.28 41 0.11 38,444
2005 445 1.13 41 0.10 39,252
2006 441 1.14 33 0.09 38,648
2007 439 1.17 20 0.05 37,435
2008 439 1.28 21 0.06 34,172
2009 327 1.06 19 0.06 30,862
2010 322 1.06 21 0.07 30,296
2011 351 1.18 20 0.07 29,867
2012 385 1.25 13 0.04 30,800
Data: Fatality Analysis Reporting System (FARS), National Highway Traffic Safety
Administration, 1990-2012 Final Files.
Note: The coding of FARS variable WEATHER was changed in 2007. Until 2006, fog was a
separate category of WEATHER. In 2007, the coding was modified so that the same
category included fog, smog, and smoke. Thus, for years 2007-2012, crashes coded above
as involving fog may have involved smog or smoke rather than fog.

15
Table 1b
All Police-Reported Crashes by Police-Reported Fog Status, United States, 1990-2008
Fog Alone Fog + Other Precipitation All Crashes
n (crashes) % of all crashes n (crashes) % of all crashes N
1990 46,264 0.72 1,987 0.03 6,462,126
1991 33,762 0.55 2,248 0.04 6,109,931
1992 31,627 0.53 1,986 0.03 5,992,938
1993 32,191 0.53 3,138 0.05 6,094,772
1994 31,884 0.49 3,834 0.06 6,489,122
1995 31,487 0.47 3,134 0.05 6,690,061
1996 32,708 0.48 2,659 0.04 6,761,051
1997 21,727 0.33 2,143 0.03 6,611,906
1998 24,651 0.39 1,840 0.03 6,325,242
1999 35,079 0.56 2,301 0.04 6,271,524
2000 31,577 0.49 6,164 0.10 6,389,310
2001 29,538 0.47 2,924 0.05 6,314,117
2002 18,317 0.29 2,212 0.04 6,304,493
2003 20,621 0.33 2,576 0.04 6,317,752
2004 20,143 0.33 3,375 0.05 6,169,998
2005 22,961 0.37 2,617 0.04 6,146,907
2006 23,583 0.40 2,878 0.05 5,964,194
2007 17,678 0.29 2,947 0.05 6,015,938
2008 20,304 0.35 2,865 0.05 5,801,228
Data: General Estimates System (NASS GES), National Highway Traffic Safety Administration, 1990-
2008.Annual Report File. Data are weighted to reflect all police-reported crashes nationwide.

16
Table 2
Characteristics of Police-Reported Fatal Crashes by Fog Status, United States,
1990-2012
Any Fog All Crashes
n (crashes) % of all crashes N
Month of Crash
Dec-Feb 4,530 2.44 185,926
Mar-May 2,036 1.02 199,789
Jun-Aug 1,891 0.83 227,681
Sept-Nov 3,396 1.55 218,791

Day of Crash
Mon-Fri 7,553 1.39 544,387
Sat-Sun 4,334 1.51 287,611

Time of crash
Midnight-5:59am 5,051 2.91 173,836
6:00-11:59am 3,662 2.31 158,768
12:00-5:59pm 598 0.25 242,555
6:00-11:59pm 2,054 0.89 230,891

Number of Vehicles Involved in Crash

1 6,753 1.40 482,201
2 4,277 1.44 296,882
3-5 758 1.47 51,586
6-9 57 4.37 1,303
10+ 42 19.53 215

Roadway Alignment (1990-2009)

Straight 6,965 1.27 548,174
Curve 2,883 1.52 189,377

Land Use (1990-2009)

Rural 7,435 1.76 423,474
Urban 2,411 0.77 314,313

Intersection-Related* (1990-2009)
Non-Intersection Non-Interchange 6,920 1.38 501,859
Intersection or Intersection-Related 2,028 1.17 173,691
Interchange or Interchange-Related 231 1.13 20,451

17
 Roadway Type (1990-2009)
Divided with Median Barrier 584 0.98 59,309
Divided No Median Barrier 1,898 1.09 174,085
Not Divided 7,190 1.49 483,455
One-Way 43 0.50 8,639

Work Zone
Yes 225 1.35 16,720
No 11,662 1.43 815,414

School zone (1995-2009)

Yes 2 1.16 173
No 7,045 1.26 557,237

Highest Driver BAC in Crash

0 6,880 1.33 517,746
.01-.07 658 1.43 45,896
.08+ 4,314 1.62 266,209

Speed-Related Crash (per NHTSA definition;

1990-2009)
Yes 3,463 1.52 228,398
No 6,422 1.25 512,761

Driver Sex** (1990-2009)

Male 4,311 1.35 320,426
Female 1,099 1.17 94,257

Driver Age** (1990-2009)

<20 862 1.46 59,043
20-29 1,887 1.54 122,614
30-39 1,086 1.32 82,026
40-49 736 1.17 62,985
50-59 462 1.18 39,310
60-69 215 0.93 23,235
70+ 156 0.65 23,918

Type of vehicle** (1990-2009)

Car / SUV / Pickup / Van 4,846 1.34 360,854
Large Truck / Bus 338 1.53 22,091
Motorcycle 201 0.71 28,355
Data: Fatality Analysis Reporting System (FARS), National Highway Traffic Safety Administration,
1990-2012 Final Files, Except where noted (2009 Annual Report File when 2009 is final year in span).

18
Note: The coding of FARS variable WEATHER was changed in 2007. Until 2006, fog was a
separate category of WEATHER. In 2007, the coding was modified so that the same category
included fog, smog, and smoke. Thus, for years 2007-2009, crashes coded above as involving
fog may have involved smog or smoke rather than fog.

* Years 1991 and later only, due to change in coding of variable.

** Includes only drivers involved in single-vehicle crashes and drivers classified as striking
(vs. struck) in multi-vehicle crashes. Multiple-vehicle crashes in which more than one driver
was classified as striking were excluded.

19
Table 3
Proportion of Fatal Crashes Involving Fog, by State/NHTSA Region, United
States, 1990-2012
Any Fog All Crashes
n (crashes) % of all crashes N
REGION 1
Connecticut 104 1.60 6,494
Maine 85 2.22 3,830
Massachusetts 125 1.33 9,426
New Hampshire 41 1.49 2,750
Rhode Island 39 2.34 1,669
Vermont 42 2.38 1,764
TOTAL REGION 1 436 1.68% 25,933
REGION 2 (excluding Virgin Islands, Puerto Rico)
New Jersey 137 0.89 15,402
New York 340 1.05 32,526
Pennsylvania 606 1.93 31,410
TOTAL REGION 2 1083 1.37% 79,338
REGION 3
Delaware 39 1.54 2,531
District of Columbia 14 1.38 1,015
Kentucky 337 1.93 17,432
Maryland 116 0.89 12,976
North Carolina 432 1.43 30,250
Virginia 183 0.97 18,826
West Virginia 269 3.33 8,077
TOTAL REGION 3 1390 1.52% 91,107
REGION 4
Alabama 269 1.24 21,626
Florida 847 1.42 59,624
Georgia 457 1.48 30,931
South Carolina 151 0.76 19,807
Tennessee 419 1.71 24,559
TOTAL REGION 4 2143 1.37% 156,547
REGION 5
Illinois 443 1.62 27,333
Indiana 296 1.60 18,517
Michigan 376 1.44 26,052
Minnesota 215 1.92 11,196
Ohio 339 1.22 27,705
Wisconsin 370 2.50 14,789
TOTAL REGION 5 2039 1.62% 125,592

20
 REGION 6 (excluding Indian Nations)
Louisiana 400 2.19 18,295
Mississippi 210 1.28 16,462
New Mexico 36 0.41 8,860
Oklahoma 154 1.07 14,361
Texas 1,062 1.53 69,521
TOTAL REGION 6 1862 1.46% 127,499
REGION 7
Arkansas 180 1.43 12,564
Iowa 195 2.19 8,890
Kansas 130 1.44 9,057
Missouri 386 1.79 21,534
Nebraska 88 1.69 5,211
TOTAL REGION 7 979 1.71% 57,256
REGION 8
Colorado 86 0.71 12,064
Nevada 20 0.30 6,732
North Dakota 41 1.89 2,164
South Dakota 80 2.51 3,190
Utah 42 0.70 5,983
Wyoming 33 1.10 3,005
TOTAL REGION 8 302 0.91% 33,138
REGION 9 (excluding Pacific Territories)
Arizona 25 0.13 19,724
California 1,001 1.25 79,864
Hawaii 7 0.26 2,729
TOTAL REGION 9 1033 1.01% 102,317
REGION 10
Alaska 22 1.29 1,709
Idaho 53 1.06 5,020
Montana 40 0.87 4,604
Oregon 256 2.71 9,433
Washington 249 1.96 12,694
TOTAL REGION 10 620 1.85% 33,460
Data: Fatality Analysis Reporting System (FARS), National Highway Traffic
Safety Administration, 1990-2012 Final Files.
Note: The coding of FARS variable WEATHER was changed in 2007. Until
2006, fog was a separate category of WEATHER. In 2007, the coding was
modified so that the same category included fog, smog, and smoke. Thus, for
years 2007-2009, crashes coded above as involving fog may have involved
smog or smoke rather than fog.

21
References

1. Rene Stutzman and Susan Jacobson. “Florida’s deadly pileup: Death toll raised to 11
as new victim found in truck.” The Orlando Sentinel, January 31, 2012. Available:
http://articles.orlandosentinel.com/2012-01-31/news/os-florida-highway-deaths-
killed-i-75-20120130_1_deadly-pileup-smoke-and-fog-first-crash; Accessed May 31,
2012
2. Deborah Whistler. “Report says I-75 opened in error before crashes that killed 11.”
Fleet Owner, May 8, 2012. Available: http://fleetowner.com/safety/report-says-i-75-
opened-error-crashes-killed-11; Accessed May 31, 2012
3. U.S. Department of Commerce, National Oceanographic and Atmospheric
Administration. Federal Meteorological Handbook No. 1: Surface Weather
Observations and Reports. (September 2005). Washington, DC, FCM-H1-2005.
Available: http://www.ofcm.gov/fmh-1/pdf/FMH1.pdf
4. Ni, Rui, Julie Kang, and George Andersen. “Age-related declines in car following
performance under simulated fog conditions.” Accident Analysis and Prevention, 42
(2010): 818 – 826
5. Broughton, Kathy, Fred Switzer, and Don Scott. “Car following decisions under
three visibility conditions and two speeds tested with a driving simulator.” Accident
Analysis and Prevention, 39 (2007): 106 – 116
6. Cavallo, V., M. Colomb, and J. Doré. “Distance perception of vehicle rear lights in
fog.” Human Factors, 43(3)(Fall 2001): 442 – 451
7. Brooks, Johnell, Matthew Crisler, Nathan Klein, Richard Goodenough, Rebekkah
Beeco, Chris Guirl, Peg Tyler, Anna Hilpert, Yarbough Miller, Jason Grygier,
Brooke Burroughs, Ashley Martin, Rob Ray, Cody Palmer, and Christine Beck.
“Speed choice and driving performance in simulated foggy conditions.” Accident
Analysis and Prevention, 43 (2011): 698 – 705
8. Mueller, Alexandra, and Lana Trick. “Driving in fog: The effects of driving
experience and visibility on speed compensation and hazard avoidance.” Accident
Analysis and Prevention, 48 (2012): 472 – 479
9. Abdel-Aty, Mohamed, Al-Ahad Ekram, Helai Huang, and Keechoo Choi. “A study on
crashes related to visibility obstruction due to fog and smoke.” Accident Analysis and
Prevention, 43 (2011): 1730-1737
10. Hewitt, Kristen. “Fog Facts.” Farmers’ Almanac (November, 2006). Available at:
http://www.farmersalmanac.com/weather/2006/11/10/fog-facts/; accessed 8/24/2012.
11. Li, Guohua, Elisa Braver, and Li-Hui Chen. “Fragility versus excessive crash
involvement as determinants of high death rates per vehicle-mile of travel among
older drivers.” Accident Analysis and Prevention, 35 (2003): 227-235
12. Centers for Disease Control and Prevention. Older Adult Drivers: Get the Facts.
(Last updated March 14, 2011). National Center for Injury Prevention and Control,
CDC. [Cited August 28, 2012]. Available:
http://www.cdc.gov/Motorvehiclesafety/Older_Adult_Drivers/adult-
drivers_factsheet.html
13. Goodwin, Arthur, Robert Foss, Lewis Margolis, and Martha Waller. Parents, Teens,
and the Learner Stage of Graduated Driver Licensing. (October 2010). Prepared for
and published by AAA Foundation for Traffic Safety. Available:
http://www.aaafoundation.org/pdf/ParentsTeensReport.pdf

22
14. Hawkins, R.K. “Motorway traffic behavior in reduced visibility conditions.” (1988).
Cited in: Brooks et al., (2011)
15. Buchner, Axel, Martin Brandt, Raoul Bell, and Judith Weise. “Car Backlight
Position and Fog Density Bias Observer-Car Distance Estimates and Time-to-
Collision Judgments.” Human Factors, 48 (2006): 300-317
16. Forthun, Gloria, Michael Johnson, William Schmitz, and Jim Blume. “Trends in Fog
Frequency and Duration in the Southeast United States.” Physical Geography, 27:3
(2006): 206-222
17. Mehler, Bruce, Bryan Reimer, Marin Lavalliere, Jonathan Dobres, and Joseph F.
Coughlin. Evaluating Technologies Relevant to the Enhancement of Driver Safety.
(2014). Washington, DC: AAA Foundation for Traffic Safety.

23